Zhuang Liu's Homepage (original) (raw)

Home
Group
News
Publications
Others

Zhuang Liu

I'm an Assistant Professor of Computer Science at Princeton University. I received my Ph.D. in Computer Science from UC Berkeley, advised by Trevor Darrell, and B.S. in Computer Science from Yao Class, Tsinghua University. Before joining Princeton, I worked as a Research Scientist at Meta FAIR, New York. I also worked as a research intern at Cornell, Intel, Adobe, and FAIR.

My research areas are deep learning and computer vision, with an emphasis on understanding how models work and behave. My work spans vision and language, unified by a focus on deep learning methods, representations, and architectures.

I explore simple approaches to gain empirical insights into neural networks. My research often challenges existing beliefs, e.g., inarch itectures,training,pruning, anddata sets.

I led the development of DenseNet (CVPR Best Paper Award) and ConvNeXt.

Research intern / visiting positions: please fill out this form to get in touch.

Research Group

PhD Students

Recent and selected publications(* equal contribution)

Transformers without Normalization

Jiachen Zhu, Xinlei Chen, Kaiming He, Yann LeCun, Zhuang Liu

CVPR 2025

[Paper] [Code] [Project Page]

MetaMorph: Multimodal Understanding and Generation via Instruction Tuning

Shengbang Tong, David Fan, Jiachen Zhu, Yunyang Xiong, Xinlei Chen, Koustuv Sinha, Michael Rabbat,
Yann LeCun, Saining Xie, Zhuang Liu

arXiv 2024

[Paper] [Project Page]

A Decade's Battle on Dataset Bias: Are We There Yet?

Zhuang Liu, Kaiming He

ICLR 2025

Zhuang Liu's Homepage (original) (raw)

Research Group

Recent and selected publications(* equal contribution)

Transformers without Normalization

MetaMorph: Multimodal Understanding and Generation via Instruction Tuning

A Decade's Battle on Dataset Bias: Are We There Yet?

Deconstructing Denoising Diffusion Models for Self-Supervised Learning

Massive Activations in Large Language Models

Eyes Wide Shut? Exploring the Visual Shortcomings of Multimodal LLMs

A Simple and Effective Pruning Approach for Large Language Models

A ConvNet for the 2020s

Rethinking the Value of Network Pruning

Learning Efficient Convolutional Networks through Network Slimming

Densely Connected Convolutional Networks

Publications(* equal contribution, for full list please see Google Scholar)

A Coefficient Makes SVRG Effective

Neural Network Diffusion

Initializing Models with Larger Ones

ConvNet vs Transformer, Supervised vs CLIP: Beyond ImageNet Accuracy

Dropout Reduces Underfitting

ImageBind: One Embedding Space To Bind Them All

ConvNeXt V2: Co-designing and Scaling ConvNets with Masked Autoencoders

Exploring Simple and Transferable Recognition-Aware Image Processing

Anytime Dense Prediction with Confidence Adaptivity

Vision Transformer Slimming: Multi-Dimension Searching in Continuous Optimization Space

Un-Mix: Rethinking Image Mixtures for Unsupervised Visual Representation Learning

Meta-Baseline: Exploring Simple Meta-Learning for Few-Shot Learning

SensAI: Fast ConvNets Serving on Live Data via Class Parallelism

Regularization Matters in Policy Optimization - An Empirical Study on Continuous Control

Test-Time Training for Out-of-Distribution Generalization

MSeg: A Composite Dataset for Multi-domain Semantic Segmentation

Few Sample Knowledge Distillation for Efficient Network Compression

Few-shot Object Detection via Feature Reweighting

DSOD: Learning Deeply Supervised Object Detectors from Scratch

Deep Networks with Stochastic Depth

Others